Introduction Low job control is associated with poor health outcomes. However, establishing the causal nature of this association is challenging mostly due to methodological biases across studies.

Objectives To examine whether changes in job control over time were associated with changes in perceptions of general health, using data from a nationally representative longitudinal study of Australian households.

Methods Our sample included 105,017 observations (18,574 persons) over 13 years from working age participants with information on job control, general health, and other sociodemographic and health factors. We used dynamic fixed-effects regression models to control for time-invariant confounding and reverse causation, and ordinary least squares regression with cumulative job control exposure to analyse cumulative exposure to low job control. Perceptions of general health, scores ranging from 0–100 points (higher scores indicating better health), were modelled in relation to self-reported job control.

Results There was a strong stepwise relationship between increasing job control and general health, with a 2.4-point increase in general health (95% CI 2.0–2.8) for participants in the highest job control quintile compared to those in the lowest quintile. The inclusion of a lagged effect of job control supported a small prospective relationship (β=0.17; 95% CI 0.07–0.27) but the contemporaneous association was essentially unchanged. Going from the lowest to highest observed value of cumulative job control predicted an 8-point higher general health score supporting a prospective association between job control and general health over a period of 4 years (β=0.47; 95% CI 0.38–0.57).

Conclusion This analysis with improved causal inference over previous research showed that increased job control is strongly associated with increasing general health. The validation of job control as a potentially modifiable causal factor of general health has considerable public health implications, and substantial economic consequences. Appropriate work stress interventions might contribute to better health and well-being.